Biodiversity in agroecosystem performs a variety of ecological services of food production, nutrient cycling, soil conservation, regulation of microclimates and water holding capacities, detoxification of pesticides and suppression of pests and diseases (Altieri, 1999). However intensive farming has affected agricultural biodiversity by toxic effects from pesticides and disturbances of the habitat structure as well as chemical conditions through fertilizer and machinery inputs as well, inflicting sustainability of agriculture (House and Alzugaray, 1989;Dennis et al., 1997;Cole et al., 2005;Uchida and Ushimaru, 2014;Vergnes et al., 2014). Unsprayed apple orchards support a large fauna of arthropod species than sprayed orchards (Bleicher et al., 2006;Kondorosy et al., 2010). Application of pesticides adversely affects ground-dwelling arthropod predators by residues, direct contact, and ingestion of contaminated prey or by feeding on dead contaminated weeds (Brown et al., 1983;Chiverton, 1984).

In apple orchards where perennial fruit trees are persistent, ground covers are often considered as important biodiversity bank. Ground vegetation can contribute biodiversity by improving soil fertility through the incorporation and retention of nutrients and water, or by hosting populations of beneficial as well as phytophagous arthropods that can provide significant stability to the agroecosystem. Ground-dwelling arthropods reproduce and mature quickly and are highly sensitivity to changes in microhabitats (Intachat et al., 1997;Ellis et al., 2001). Many are acting decomposers, arthropod predators or seed predators (Stinner and House, 1990;Ball and Bousquet, 2001;Tooley and Brust, 2002;Torres and Ruberson, 2007). Also, some members are involved in soil ecological engineering through decomposition, maintaining soil structure, fertility, and nutrient cycling (Seastedt and Crossley, 1984).

In an orchard system, weed mortality is also governed by seed predation along with chemical control practices (Menalled et al., 2000;Davis et al., 2003;Gallandt et al., 2005;Westerman et al., 2005;Menalled et al., 2007). These alterations may cause direct changes in microclimatic conditions, soil properties, and resource availability (Li et al., 2009).

Early season weed management can help apple tree for competing nutrients and water with ground vegetation (Merwin and Stiles, 1994;Utkhede and Hogue, 1998). In organic farms, physical and cultural weed control techniques are applied (Ock and Pyon, 2011), while clean management, especially desiccating weeds by herbicide or complete bare-ground maintenance is practiced in conventional farming (Choi et al., 2009).

In this study, we monitored ground-dwelling arthropods using pitfall traps in organic and conventional apple orchards. In agriculture ecosystems, Hymenoptera, Coleoptera, and Araneae are dominant ground-dwelling arthropods (Purvis and Curry, 1984;Bhatti et al., 2005). In Coleoptera, Carabid beetles are free-living generalist predators and considered as ground cover indicators because of their sensitivity to habitat alteration (Holliday and Hagley, 1984;Barney and Pass, 1986;Andersen, 1992;Cárcamo, 1995;Cárcamo et al., 1995;Kromp, 1999). Thus, objectives were to evaluate if the ground-dwelling arthropod assemblages, mainly Hymenoptera, Coleoptera, and Araneae are more influenced by farming systems or weed management.

Materials and Methods

Study sites

The study was conducted in 10 apple orchards at five different locations as follows: Buksam in Chilgok (36°03´N, 128°12´E); Dain (36°27′N, 128°23′E) and Oksan (36°21′N, 128°52′E) in Uiseong; Hyeondong in Cheongsong (36°17′N, 129°00′E); Punggi in Yeongju (36°50′N, 128°28′E) in Gyeongbuk province, South Korea (Fig. 1; Table 1). In each location, a pair of organic and conventional apple orchards were chosen within 500 m distance each other. 3 ~ 9-year-old organic orchards were chosen for this study, and “Fuji” was the most dominant apple variety in study sites (Table 1). Weed managements of the two orchard systems are shown in Table 1. In organic farms, weeds (ground vegetation) were remained or trimmed with mowing machine between the rows of trees, while in conventional orchards, herbicide spay or bare-ground maintenance is practiced, in general. Likelihood rate of practices was depicted in Table. 1. If mowing is practiced every year in spring, then “1”, or once in 3 year, then “0.3”.

Sampling of ground-dwelling arthropods

Using pitfall traps (8 cm in diameter of mouthpart, 12 cm in height) ground-dwelling arthropods were collected in the spring (late April-mid May) from each site every year during 2011-2013. Ethylene glycol (50 ml per each trap) was used as preservatives for preservation of collected arthropods. Ten pitfall traps were placed in each site, and the traps were replaced by 3 days interval for 9 days in each site. After collection the arthropod samples were classified into the higher taxonomic level (e.g., Order). Coleopterans were further classified into family level, and then carabid beetles were identified into species level based on morphological characteristics.

Statistical analysis

Data obtained for ground-dwelling arthropod abundances between organic and conventional orchards were normalized with a log (N+1) transformation after Shapiro-Wilk test (Macfadyen, 1962). Abundances of functional groups of ground- dwelling arthropods and diversity of carabid beetles between organic and conventional orchards were analyzed by using split plots analysis of variance (ANOVA). For split plots ANOVA, year and location were treated as the split factors. Means were separated by Tukey’s HSD test for all parameters (SAS Institute, 2010). Composition and diversity of carabid beetle communities were assessed by species richness (Kessler et al., 2001), Shannon-Wiener diversity (Shannon and Weaver, 1949), evenness (Pielou, 1966), and dominant species. The dominant species category was selected based on an abundance greater than 5% of total (Brockmann-Jerosch, 1907). Pearson correlation was used to examine relationships between number of individuals in major ground-dwelling arthropods and number of weed managements across all three years. Further, nonmetric multidimensional scaling (NMS) was conducted to analysis the composition of ground-dwelling arthropods based on Euclidian distance matrix among study sites and study years (PC-ORD ver. 6, MJM Software Design, Gleneden Beach, OR). Multi-response permutation procedure (MRPP) was used to determine the distinctness of ground-dwelling arthropod communities between the different weed managements (McCune and Grace, 2002).

Results

Arthropod taxonomic and functional groups

In total 2,825 individuals were collected from 10 orchards for 3 years, and identified as 73 species belonging to 33 families (13 orders). In organic orchards, the 1,659 collected individuals comprised 59 identified species, while in the conventional orchard the 1,166 collected individuals consisted of 47 species. The community structure of ground-dwelling arthropods in two farming systems is presented in Table 2. Similar proportions of arthropod groups were found in the two apple orchard systems and involved Coleoptera, Araneae, Hymenoptera, Diptera, and other arthropods, accounting for 41.5, 38.0, 12.7, 3.8 and 3.9%, and 44.3, 36.5, 13.2, 3.5 and 2.5% in the organic and conventional orchard, respectively (Fig. 2A). In Coleoptera, the most commonly encountered group were the carabids, accounting for 86.0% (621 individuals belonging to 23 species) and 83.5% (431 individuals belonging to 16 species) in organic and conventional orchard, respectively (Fig. 2B).

Abundance of ground-dwelling arthropods was significantly different between farming systems (F = 7.52, df = 1, 16, P = 0.01), among locations (F = 6.19, df = 4, 16, P= 0.003) and among years as well (F = 8.06, df = 2, 16, P= 0.003) (Table 3). In functional groups, the abundances of herbivores (F= 8.55, df = 1, 16, P = 0.009) and predators (F = 10.21, df = 1, 16, P = 0.005) were significantly differed between farming systems but parasite abundance exhibited no difference (Table 3; Fig. 3). By considering locations and years, predators were significantly different among locations (F = 3.68, df = 4, 16, P = 0.02) and years (F = 24.57, df = 2, 16, P< 0.0001), but those factors did not different in herbivore abundances (Table 3). Parasite abundances were only significantly different in connection with different years (F = 5.95, df = 2, 16, P = 0.01)(Table 3). Significant interaction between farming system and location was found for abundance of ground-dwelling arthropods and predators.

Carabid beetle diversity

Species richness of carabid beetles was significantly different between farming systems (F = 16.88, df = 1, 16, P = 0.0008) and among years (F = 11.38, df = 2, 16, P = 0.0008) (Table 4). Abundance was also significantly different between farming systems (F = 12.40, df = 1, 16, P = 0.002), among locations (F = 6.30, df = 4, 16, P = 0.03), and interaction as well. Diversity and evenness were significantly different between farming systems (Diversity: F = 6. 54, df = 1, 16, P = 0. 02; Evenness: F = 15.18, df = 1, 16, P = 0.001) (Table 4).

Both organic and conventional apple orchards, Anisodactylus punctatipennis, Amara congrua, Nebria coreica, Harpalus discrepans, Anisodactylus sp. and Chlaenius pallipes were dominant (Table 5).

Influence of weed management on arthropods

The correlation between abundance of ground-dwelling arthropods and number of weed managements (herbicides and mowing) is presented in Table 6. Abundances of Araneae, Coleoptera and carabids were negatively correlated with the weed management frequencies. Furthermore, the presence of diplopod, Nematophora was negatively correlated with mowing.

Non-metric multidimensional scaling showed overlaps among different weed managements in organic and conventional apple orchards (Fig. 4). In this NMS ordination, axis 1 and axis 2 were explained by 45 and 47% of the variance, respectively. MRPP confirmed that the composition of ground-dwelling arthropods between two weed management practices was significantly different (A = 0.04, P < 0.05).

Discussion

This study confirmed that the abundance of total grounddwelling arthropod is highly variable depending on farming system, locations of orchards and even among years. This imply that the ground dwelling arthropod assemblages are not stable but vulnerable to environmental changes (Intachat et al., 1997;Ellis et al., 2001). In functional guild analysis, abundance of predator group was the highest compare to other groups, such as detritivores, herbivores, or parasites, which indicate that the majority of ground arthropods play roles in predation and could be beneficial as natural suppressors of insect pests (Marc et al., 1999). This is partly due to the higher abundance of spiders and carabid beetles (Kromp, 1999). Other results that the herbivore and predator groups were significantly high in organic apple orchards than in conventional ones may indicate the complex biological diversity and food web structures in organic orchards where minimum chemical perturbations are applied. Hasin and Booncher (2020) reported larger occurrence of rove beetles and spider in organic farms than in conventional ones. Similar results applied to herbivores including primarily Acrididae, Membracidae, Melolonthidae, and Cetoniidae. Many herbivore groups are not only grass feeders but also deciduous tree foliage, pollen and nectar feeders. Since organic apple orchards are less disturbed by chemical pesticides, large sized herbivores could escape from pesticide spray and invade into apple orchards (Lee et al., 1996). The abundance of detritivores did not differ between the two orchard systems. In another study (Kim, 2016), detritivore abundance and richness were primarily governed by the organic input into the orchards such as in organic straw mulching or chopped bark. In this studied orchards, organic input was not systematically differed between two farming system. Also, even limited dispersal capacities, detritivores are often influenced by the large and complex landscape context rather than on-farm management (Flohre et al., 2011).

Abundance and diversity of carabid beetles in organic and conventional orchards are linked to weed management practices, physical conditions of ground such as temperature and humidity, types of groundcover, and food resources as well. In general, carabid beetle adults live on the soil surface and are highly mobile, while larvae live in the soil, and possess limited dispersal capacity (Kromp, 1999). In this aspect, as underground weeds in orchards provide valuable ecological services such as moderating the spatio-temporal habitat conditions and providing herbivores or seeds as foods for grounddwelling arthropods including predators, and eventually supports biological diversity within cultivated fields (García-Ruiz et al., 2018). For example, species belonging to Amara, Anisodactylus and Harpalus are known to prefer habitats with increased humidity due to their suitability for oviposition or larval development (Holland, 2002), especially in regard to species that as larvae and adults rely on seeds for food (Luff, 2002). The similarities of the fauna might be related to the movement of carabid beetles from overwintering sites of surrounding habitats in spring season and their longevities (Willin, 1985;Coombes and Sotherton, 1986;Purtauf et al., 2005). It is likely that the abundance of carabid beetles is related to ground vegetation maintenance. Now in May orchard system, therefore, ground weeds are needed intentionally management of underground vegetation, such as companion plants (Kishinevsky et al., 2017).

Overall, ground-dwelling arthropods taxonomically and functionally were significantly affected by farming systems as well as weed management frequencies in this study. In organic farming, the weed management frequency is one of important factors in the farming systems, because the removing ground vegetation more altered the ground habitat structure drastically compared to other the organic farming, which is more focused on foliage protection with restriction of chemical pesticides and fertilizers. It is clear that even though organic farming could counteract the detrimental effects of agricultural intensification on farmland biodiversity for sustainability, functional diversity of arthropod could be better achieved through intentional engagement of habitat manipulation using companion plants to facilitate the preemptive occupation of agricultural ecosystem.